Yep. I'm aware of this side effect. The question is whether or not this is an issue. It basically is going to come down to defaults; that is, if no direction is specified, and the direction cannot be determined from some other means, can I safely assume a default value of "ltr".. in which case, I can specify dir="rtl" at the feed level, and throw in a dir="ltr" on entries that otherwise did not have any directionality specified. If that is an acceptable behavior (which I think it might be) then there is no problem dropping the empty string value. If it is not an acceptable behavior, then we need the empty string value.
- James James Holderness wrote: > > James M Snell wrote: >> [EMAIL PROTECTED] wrote: >>>> From the example, I am not sure what the feed author wants. So far >>>> there are only two directionalities allowed, left-to-right and >>>> right-to-left. In the example, the directionality of atom:entry is >>>> "rtl", there is no reason for the author to reset the directionality >>>> with an empty string. He should either change the directionality to >>>> "ltr" in atom:content or keep "rtl" by not using any "dir" attribute >>>> in atom:content. >> >> Fair enough. Removing dir="" requires that the feed publisher always >> explicitly set the direction. It simplifies processing and helps to >> eliminate guessing. Those are compelling enough reasons. In the next >> iteration I'll remove the blank value. > > What would a planet feed do when aggregating content from other feeds > that may not have explicit directionality? If I'm understanding things > correctly, you would lose the ability to use the dir attribute on the > root of the feed, because there would be no way of overriding it for > feed items whose directionality was unknown. > > Not that that's the end of the world. I just want to be sure that you > realise you'd be losing that functionality. > > Regards > James > >
